A suitable technique to upgrade biomass is an essential point in order to make it possible to develop a sustainable economy based on biomass sources. One likely and green option to extract chemicals from biomass is hydrothermal processing. A process that only requires water, high temperature and pressure to fractionate biomass. In the same way, understanding what this fractionation really is like, is an essential factor to design industrial scale biorefineries too. Thereby, a comprehensive modelling of this kind of processes must be done, which would enable us to go from a lab-scale set up to an industrial process. As a result, this work is aimed at developing a model able to predict solubilized biomass hydrolysis. Holm oak, as biomass sample, was fractionated in a packed bed reactor at 180 ºC, producing a solid residue and a liquid stream. This latter stream was hydrolyzed afterwards by water in a tubular reactor whose operational conditions ranged from 350 ºC to 396 ºC.
The model was focused on this last reactor, applying a transient mass balance for each compound produced (oligomer, sugar and degradation product concentration). The result was a set of 4 partial differential equations that was discretized by orthogonal collocation on finite elements method and the obtained ordinary differential equations were solved by 8th convergence order Runge-Kutta’s method. The adjustment was performed by a modification of the SIMPLEX method and the average deviations were lower than 10 % in all the cases. Similarly, the total organic carbon profile was also successfully reproduced with no discrepancy between the simulated and the experimental values. Moreover, not only did it reproduce the experimental behavior but also it successfully simulated the evolution of the hydrolysis kinetics along time.
This study has been supported by the professor training Spanish program (FPU2013/01516) and the Spanish projects: BioFraHynery CTQ2015-64892-R and VA330U13